Kimahriman commented on code in PR #52642:
URL: https://github.com/apache/spark/pull/52642#discussion_r2475679774
##########
sql/core/src/main/scala/org/apache/spark/sql/execution/streaming/runtime/MicroBatchExecution.scala:
##########
@@ -344,9 +344,40 @@ class MicroBatchExecution(
setLatestExecutionContext(execCtx)
populateStartOffsets(execCtx, sparkSessionForStream)
+
+ // SPARK-53941: This code path is executed for the first batch, regardless
of whether it's a
+ // fresh new run or restart.
+ disableAQESupportInStatelessIfUnappropriated(sparkSessionForStream)
Review Comment:
Do you have any plans to try to either remove this or have a way to opt-out
since other checks _should_ theoretically be enough when choosing wither to
insert an adaptive plan? We (sadly) only recently discovered the performance
benefits of manually re-enabling AQE inside a FEB and it would be nice to have
that work by default, and I think this is still preventing that from happening?
--
This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
To unsubscribe, e-mail: [email protected]
For queries about this service, please contact Infrastructure at:
[email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
